Generally, in order to embody a light device such as a light emitting diode or a laser, a good ohmic contact should be first of all made between a semiconductor and a metal formed as an electrode.
Further, a flatted surface state, a thermal stability, easy processing, a low contact resistance, a high yield, a good corrosion resistance and the like are required.
In the meantime, a GaN-based nitride semiconductor light emitting device is mainly grown-up on a substrate 2000. The substrate 2000 may be a sapphire substrate or a SiC substrate. Additionally, a GaN-based polycrystalline layer is grown-up as a buffer layer on the sapphire substrate or the SiC substrate at a low growing-up temperature, and then a first conductive layer 2001 (e.g., an undoped GaN layer, a silicon (Si) doped N-type GaN layer, or an N-type GaN-based layer having a combination structure thereof) is formed on the buffer layer at a high temperature. After that, a light emitting layer (quantum-well-structured active layer 2002) is formed on the first conductive layer 2001 (e.g., the N-type GaN-based layer), and a P-type GaN-based layer is additionally formed on the light emitting layer such that the semiconductor light emitting device is manufactured.
Additionally, in the semiconductor light emitting device, a transparent electrode can be formed in the following method.
First, a P-type electrode structure formed in a conventional light emitting device is briefly described with reference to FIG. 1.
FIG. 1 is a view illustrating an exemplary P-type electrode of a conventional light emitting device.
The P-type electrode of a light emitting device shown in FIG. 1 is structured to have a P-type transparent electrode layer 102 formed on a P-type GaN layer 101, and have a P-type bonding electrode 103 formed on the P-type transparent electrode layer 102. The above-structured electrode structure is called a ‘close’ electrode structure for convenience.
In case of the ‘close’ electrode structure, the P-type transparent electrode layer 102 is mainly formed of Ni/Au layer. Additionally, the P-type bonding electrode 103 is of a single layer including two or more metals (for example, Au, Ti, Ni, In and Pt) based on Au except for Al and Cr, or a multi-layer structure of two or more layers. That is, it is of an Au, Ni/Au, Ti/Au or Pt/Au layer and the like.
For example, as shown in FIG. 2, one metal is selected from the group consisting of Ni, Pt, Ti, Cr and Au to deposit a first metal layer 102a on the P-type GaN-based layer 101, and gold (Au) is used to deposit a second metal layer 102b such that the transparent electrode 102 can be formed. As a typical example of the transparent electrode, a Ni/Au electrode is used.
Or, as shown in FIG. 3, a first metal layer 102c from which oxide is well-formed is formed on the P-type GaN-based layer 101 and then, after a second metal layer 102d for carrier conduction, for example, Gold (Au) is deposited, a thermal annealing is performed in an oxygen-containing atmosphere.
As a typical example, there is a method in which after cobalt (Co) and gold (Au) are sequentially deposited on the P-type GaN-based layer 101, the thermal annealing is performed in the oxygen-containing atmosphere to form a ‘Co—O’ oxide. Or, a method is also proposed using nickel (Ni) instead of cobalt (Co).
Accordingly, a metal oxide layer 102e is formed to have transparency such that the transparent electrode 102 is formed on the P-type GaN-based layer 101.
A conventional P-type electrode of the light emitting device can be also structured as shown in FIG. 4, and FIG. 4 is a view illustrating another exemplary P-type electrode of the conventional light emitting device.
The P-type electrode of the light emitting device shown in FIG. 4 is structured to have a P-type transparent electrode layer 202 formed on a P-type GaN layer 201 and have a P-type bonding electrode 203 formed on the P-type transparent electrode layer 202. At this time, the transparent electrode layer 202 is structured to have a portion of the P-type bonding electrode 203 filled therebetween.
The above-structured electrode structure is called an ‘open’ electrode structure for convenience.
In case of the ‘open’ electrode structure, a structure including a Cr or Al layer is proposed so as to improve a bonding capability, and is formed to have a similar structure with the ‘close’ electrode structure.
In the meantime, FIG. 5 is a view illustrating an exemplary N-type electrode of the conventional light emitting device.
The light emitting device shown in FIG. 5 is structured to have an N-type electrode layer 302 formed on an N-type GaN layer 301.
In case of the N-type electrode layer 302, proposed is a single layered electrode using Ti, Al, Au or a multi-layered electrode of two or more layers.
However, in case of the above structured P-type electrode, the specific contact resistance is much higher than 10-3 Ωcm2 due to a high resistive P-type GaN layer.
Further, it is known that in a transparent electrode structure (referring to FIG. 2), not an oxide structure, since the specific contact resistance is as much high as 10−2 Ωcm2 a ‘current spreader’ that is one of the most main functions of the transparent electrode does not function during the device operation.
It is known that this acts as a heat source at an interface at the time of a device operation due to a high specific contact resistance of the interface to directly cause much influence on a degradation of device reliability.
Further, it is reported that the transparent electrode structure formed by the fabrication method described with reference to FIG. 3 has the considerably improved specific contact resistance, but it is known to have a performance deteriorated in light transmittance. It is known that this is since a metal oxide is ‘polycrystalline’-structured at the time of the thermal annealing in an oxygen atmosphere, not ‘heteroepitaxial’ structure contributable to improvement of transmittance, a number of small-sized grains existing within the transparent electrode cause absorption or scattering loss of photons emitting from the semiconductor.
Furthermore, in order to embody a good quality ohmic electrode in the above structure, the carrier should have a concentration of more than 1018 cm−3 in a doping region in which a carrier tunneling can be made, but actually a carrier concentration of a P-type Gallium nitride-based compound semiconductor is as much low as 1017 cm−3.
As such, the low carrier concentration causes a Schottky barrier height to increase the specific contact resistance at the interface between the metal and the semiconductor, resulting in the poor ohmic characteristics.
Further, a native oxide layer existing a surface of a P-type gallium nitride-based compound semiconductor causes a mutual reaction at the interface between the metal and the semiconductor at the time of the thermal annealing such that many drawbacks of an increase of a leakage current, a reduction of a reverse breakdown voltage, an abnormal threshold voltage characteristic and the like are caused, and as a result, the device reliability and life-time are reduced.
Furthermore, the above drawbacks occur from all of the light emitting device including the P-type electrode of the ‘open’ electrode structure and the ‘close’ electrode structure. Accordingly, the P-type electrode with a high thermal stability and a low contact resistance is so sincerely required to be developed.
Additionally, the N-type electrode having the specific contact resistance of more than 10−5 Ωcm2 is suitable to the light emitting device, but a Ti-based electrode is reportedly so vulnerable in view of a thermal characteristic.
Further, the conventional art has many disadvantages in view of a device production and yield since the P-type electrode and the N-type electrode are separately manufactured.